It is now well documented that several human retroviruses including HIV can induce central nervous system diseases with clinical manifestations ranging from subtle personality, perceptual motor and memory disorders to frank dementia. In addition, HTLV I appears to be the causative agent of tropical paraparesis a devastating motor disease encountered in Africa. It is clear that these viruses replicate in the CNS and for HIV at least, the cell types which are infected have been identified. We are studying a neurovirulent retrovirus of mice which, although only distantly related to the human retroviruses, induces a disease with manifestations which in some respects resemble those of the human diseases. Our primary goals have been to identify the important target cells in the CNS the infection of which leads to clinical disease and to determine the nature of the cytopathology that ensues upon such infection. Like HIV infection, clinical signs of neurologic disease were found to be unaccompanied by any significant histopathology. This has suggested to some investigators that infection of endothelial cells in the CNS may itself mediate the CNS dysfunction. However, we have found no correlation between replication of murine retroviruses in CNS endothelial cells and neurologic disease. Disease was clearly associated with the capacity of WM-E to spread to the parenchyma of the CNS, and, using molecular constructs between the neurovirulent and a non-neurovirulent viruses, extravascular spread was found to be mediated by the viral envelope gene.